Open to opportunities

Vikas Dubey

@vikasdubey

Message

Principal AI/ML Engineer specializing in GenAI architecture, RAG, and agentic pipelines to deliver production outcomes.

India

Message

What I'm looking for

I’m looking to lead GenAI/LLM systems end-to-end—owning architecture, evaluation, and production delivery. I want roles where I can define technical strategy, build RAG/agentic pipelines, and partner cross-functionally to drive measurable business outcomes.

I’m a Principal AI/ML Engineer with 12+ years owning end-to-end GenAI system architecture across Tech (Apple, Google), Telecom, and BFS domains. I bring deep expertise in LLM evaluation frameworks, agentic pipeline design, RAG architectures, and production ML at scale, with a track record of defining technical strategy and leading cross-functional implementation.

At Fractal Analytics, I’m the sole architect and tech lead for an LLM-powered thread resolution scoring system for Apple, delivering 94% classification accuracy across 10,000+ community threads. I’ve also architected an LLM-based insight engine for Google (1M+ YouTube comments) and led a time series WFM forecasting initiative for a US Telecom client, with production deployment under a Kedro framework and dashboards to help optimize agent schedules.

Mentored and guided junior data scientists and interns across multiple project cycles — covering LLM workflows, ML model development, and GenAI best practices; provided cross-team technical support beyond assigned project scope

Experience

Work history, roles, and key accomplishments

Current

Senior Data Scientist (GenAI Lead)

Current

Fractal Analytics

Nov 2021 - Present (4 years 8 months)

Owned end-to-end GenAI system architecture for Apple community thread resolution scoring using Gemini 2.5 Pro, delivering 94% classification accuracy across 10,000+ threads. Also architected LLM insight engines for Google (1M+ YouTube comments) and led a time-series workforce forecasting framework for a US telecom client with Kedro and Tableau dashboards.

Fbprophet GenAI LLM Architecture LLM Frameworks Prompt Engineering RAG LlamaIndex Function Calling Vector Databases Python Pyspark fastAPI SQL Kedro ARIMA GreyKite Tableau Time Series

Senior Data Scientist

Merkle Inc

Jul 2019 - Nov 2021 (2 years 4 months)

Built and productionized PySpark classification models for telecom cross-sell/up-sell targeting, with decile charts used for marketing customer targeting. Developed end-to-end media mix models in the L&D domain using R and delivered Tableau dashboards and campaign evaluation metrics.

Pyspark Classification Cross Selling Media Mix Modeling R Tableau Marketing Analytics

Data Scientist

Cemtics

May 2018 - Jun 2019 (1 year 1 month)

Delivered two telecom domain projects with multiple model variants deployed to production. Built classification and regression models using traditional ML and deep learning approaches including tree-based methods and neural networks (Keras/TensorFlow).

Classification regression Logistic Regression Decision Trees Random Forest SVM XGBoost CatBoost Deep Learning Keras TensorFlow

Associate Consultant

Capgemini

May 2017 - Dec 2017 (7 months)

Delivered banking domain ML models including EDA, statistical testing, regression and classification. Performed data integration using Oracle SQL and built models with linear, logistic, and tree-based methods.

Machine Learning EDA regression Classification Logistic Regression Decision Trees Random Forest Oracle SQL Data Integration

Project Engineer

Wipro

Dec 2013 - Apr 2017 (3 years 4 months)

Built foundational ML pipelines in the banking domain, combining data from multiple sources using Oracle SQL. Conducted EDA and statistical analysis and developed predictive models using linear, logistic, and tree-based approaches.

Machine Learning Pipelines Oracle SQL EDA Statistical Analysis predictive modeling Logistic Regression Decision Trees Random Forest